Otimização do despacho econômico de carga com a utilização de heurísticas bio-inspiradas, considerando um limite de penetração de fontes alternativas

Abstract

The increased consumption of electricity in the 21st century, is changing the consumer market, making it highly competitive, changing several aspects that affect the generation, transmission and distribution of electric energy. In the search for solutions to meet consumer demand and reduce environmental impacts, distributed generation (DG) is being widely inserted into the electric power system (EPS), directly influencing the economic load dispatch (ED). In this way, the optimization of the economic load dispatch, which is a classic subject with initial studies prior to the 1920s, has become even more relevant for network operators, since the diversification of the energy matrix has brought a difficult to the task of harmonizing electric energy sources that have very different characteristics in terms of costs, operation and investments. In this context, this work aims to present solutions obtained trough the application of optimization algorithms, using a stopping criterion and some operational restrictions. Among the restrictions, it was given highlight to those related to losses in the transmission lines, as well to the limits of the generation capacities of the thermoelectric plants and a penetration limit of the distributed generation. The algorithms used to solve the problem were, respectively, the Lagrange Multipliers Method (LMM), the Bat Algorithm (BA) and the Particle Swarm Algorithm (PSO), which were applied to the solution of two test systems, being the first with three and the second with six generating units with and without a percentage of DG insertion. In addition, the biological principles that motivated the development of the heuristic methods applied in this study are demonstrated, as well as the computational algorithms were introduced. The results were satisfactory using the mathematical method LMM as a comparative basis, since it returns the global optimum for the objective functions of each situation analyzed. Thus, it was possible to conclude that the BA obtained a better convergence to the global minimum of the objective function when compared to the performance of the PSO for the problem of six thermoelectric plants. For the problem of 3 generating units, PSO had a slight advantage over BA.